What Works and Why

A few years ago, I spent entire days manually verifying email addresses. Copy an address. Paste it into a verification tool. Wait. Copy "safe to send" back into my spreadsheet. Repeat 299 more times. My eyes would glaze over by noon and I'd still only be a fraction of the way through my list.

Fast forward to now and our stack processes 1,000+ contacts a day without breaking a sweat. The tools do in 10 minutes what used to take me all week.

But powerful tools don't automatically make you better at outbound. We had to learn what each platform was actually built for and more importantly, what to stop forcing it to do.

The real breakthrough wasn't finding the "perfect" tool. It was learning how to make them work together, each one covering the gaps in the others.

Here's what we use and why it works.

Apollo: The Friend Who Knows Everyone

What we use it for: Apollo is where everything starts. It's the backbone of our entire prospecting operation.

Why we love it: Apollo's search capabilities are the best we've found for B2B contacts. We're talking millions of contacts with filters granular enough to find exactly who we need: specific roles, company sizes, industries, tech stacks, funding stages, whatever matters for a particular campaign.

What really sets Apollo apart is how it brings everything together in one place. We can build lists, verify contact info, track engagement and manage sequences without jumping between platforms. That unified approach saves us hours every single day.

Apollo's data accuracy sits around 85-90% in our testing. That's really solid, but not perfect, which is why we layer in additional verification and enrichment (more on that below).

How we use it: Apollo is our first and often our last stop in the prospecting journey. We start by building targeted lists with their powerful filters. Then, after enriching and verifying through our other tools, we circle back to Apollo for additional enrichments on qualified leads.

It's the tool we couldn't run our agency without.

Clay: Where We Got a Little Obsessed

What we use it for: Data enrichment, workflow automation and basically anything that requires pulling information from multiple sources. Clay is where the possibilities really open up.

What's really good: Clay can pull firmographics, technographics, social profiles, funding data, you name it. Multiple data sources mean richer profiles and better personalization angles.

But what really sold us is the flexibility. We can clean lists from Apollo, build brand new ones from scratch, connect APIs from other tools and create completely custom workflows for each client. Our process with every client is different and Clay handles that without breaking.

We started using Clay just for basic enrichment (verifying emails, pulling company info, standard stuff). Then we started connecting it to other tools in our stack and realized we could automate entire research workflows that used to take hours.

Now we can personalize outreach at scale in ways that actually feel relevant to prospects. The kind of research that makes someone reply "How did you know we were dealing with this?"

The measurable impact: For one client, we used to average 30-35 qualified leads per month using basic personalization. After implementing Clay enrichments through the OpenAI and Anthropic APIs (GPT and Claude), we jumped to averaging 70-80 qualified leads per month. More than doubling results with the same outreach volume.

The brutal truth about Clay: Garbage in, garbage out. If your input data is messy (bad domains, wrong titles, broken LinkedIn links), your enrichments will be just as messy. Data cleaning isn't optional with Clay.

Also, credits can disappear fast if you're not careful. We once enriched 5,000 contacts with every data point Clay offered. Burned through $300 in credits and 90% of the "insights" never got used in our campaigns. Now we're surgical, only enriching what actually improves our messaging and reply rates.

The learning curve: Clay takes time to learn properly. You're constantly discovering new possibilities, new integrations and new ways to automate things you didn't realize could be automated. But it's worth the effort. Once it clicks, you start seeing workflows everywhere.

Smartlead: The Reliable Workhorse

What we use it for: Email sending, sequencing and campaign management. This is where our carefully researched, enriched, personalized outreach actually gets sent.

What matters most to us:

The whitelabel capabilities are crucial when you're running an agency.

Deliverability is non-negotiable. Smartlead's infrastructure keeps our emails out of spam folders, which matters more than any clever subject line. We can send volume without tanking our sender reputation.

The API lets us connect everything else in our stack. Automation only works when your tools actually talk to each other.

Unlimited mailboxes mean we can scale campaigns without hitting arbitrary limits. As client needs grow, we're not scrambling to upgrade plans.

The constant updates and upgrades keep the platform evolving. We're not stuck with a tool that stays the same while email changes around it.

The catch: What frustrates us most about Smartlead isn't the sending infrastructure, that's solid. It's the campaign management UX. Some basic functions that should be straightforward get overlooked as the team builds new features. Issues with core campaign and contact management linger longer than they should.

The interface for managing campaigns and contacts could be significantly better. When you're juggling multiple clients and dozens of campaigns, clunky UX adds friction you don't need.

Why it works for us: Smartlead doesn't try to be everything. It's focused on doing one thing well: getting emails delivered and managing responses at scale. That focus means it plays nicely with our other tools instead of trying to replace them.

DeBounce: The Insurance Policy

What we use it for: Email verification and list cleaning. Before any outreach campaign, everything runs through DeBounce.

Why it matters: Every bad email you send damages your domain reputation. High bounce rates signal to email providers that you're not maintaining your lists properly, which pushes your future emails into spam folders—even the legitimate ones.

DeBounce catches invalid emails, catch-all domains, disposable addresses and role-based emails before we ever hit send. It's the unglamorous insurance policy that protects everything downstream.

What's really good: It's cost-effective and it works. By catching bad emails before we send, we protect our domain reputation and keep our deliverability high. Most campaigns maintain bounce rates under 1%, which keeps our sender scores healthy.

The honest part: We've always used DeBounce, so we can't give you a dramatic "we learned this the hard way" story. But we see what happens to teams that skip verification, they end up on our calls asking why their emails suddenly started landing in spam. The answer is always the same: they burned their sender reputation by not cleaning their lists.

It takes weeks to rebuild a damaged domain reputation. DeBounce starts at $10/month. The math isn't complicated.

GPT + Claude: The Research-to-Copy Pipeline

Here's where AI earns its keep in our workflow, but not in the way most people think.

The workflow: We use GPT for the research heavy lifting. It digs through websites, analyzes company information, finds angles that would take forever to uncover manually. It's thorough in ways that would require an entire research team otherwise.

Then Claude helps with the copywriting. But (and this is important) it's not writing our emails for us. We use it to generate one or two sentences that add to the relevance and personalization of our emails. The human touch is still there. AI helps us work faster and sharper, but every variation gets a human review before it goes out.

The review system: We don't review every single email, that would be impossible at scale. We review every email variation. A sequence might have three steps with four variations on step one, three on step two, and one on step three. We check all variations to ensure the AI output matches our template and the spin text we're using in our copy.

But the real quality control happens earlier, during the AI enrichment phase in Clay. We have a multi-layer review process:

Keyword filters flag outputs containing words we want to avoid

AI reviews AI - we run outputs through a content safety review model

Human review for anything flagged by either system

Any output that gets flagged gets rewritten by hand. We're especially careful with cultural sensitivity, tone matching and avoiding anything that could come across as patronizing or tone-deaf.

Why this combo? We've tested a bunch of AI writing tools. Claude is the best at staying conversational without trying too hard. It doesn't sound like a robot attempting a human impression.

The key is knowing when to use AI and when to trust your own instincts. AI can accelerate research and help polish copy, but it can't replace understanding your audience or knowing what message will actually resonate.

Make: The Glue That Holds Everything Together

What we use it for: If Clay is where data gets enriched and Apollo is where it originates, Make is what connects everything and keeps it flowing seamlessly.

Why it's essential: Make lets us build automations that make our workflows faster and more efficient. Every qualified lead we generate goes through Make for distribution and client notification.

The workflow that saves us the most time:

When a positive reply comes into Smartlead, here's what happens automatically:

1. We apply a category label to the email thread (based on the reply type)

2. That label triggers a Make workflow

3. Make enriches the lead data even further by pulling additional context from Apollo

4. The lead information gets added to our Notion database

5. Simultaneously, it updates the client's Google Sheets

6. A notification goes out to the client's Slack channel

7. An email summary gets sent to the client

8. Everything syncs across all our systems

This entire process takes seconds. Without Make, we'd be manually moving data between platforms, copy-pasting information and probably missing steps in the chaos. With Make, everything flows automatically, consistently, and accurately.

What breaks if Make goes down? Lead handoff stops completely. Clients don't get notified. Data doesn't sync. We're back to manual processes until it's back up. That's how central it is to our operations.

Why it works for us: Make gives us the flexibility to build exactly the workflows we need. If we can imagine a process, we can usually build it in Make. That means as our needs evolve or as we onboard new clients with unique requirements, we're not stuck. We just adapt the automation.

Notion: Our Command Center

What we use it for: Pretty much everything that isn't directly sending emails. Task management, lead databases, client reports, internal documentation. It all lives in Notion.

What's really good: The shared page features are clutch for client collaboration. We can create dedicated spaces where clients access reports, resources and relevant docs without drowning them in email attachments or Slack threads.

It keeps everyone on the same page (literally) and makes our processes transparent. Clients can check campaign progress, review lead quality or reference strategy docs whenever they want.

Why it works for us: Notion scales with complexity without becoming overwhelming. As our processes get more sophisticated and our client base grows, we can organize everything in ways that actually make sense. Plus the flexibility means we're not fighting against rigid structures that don't fit how we actually work.

How These Tools Actually Work Together

Here's the step-by-step flow of a typical campaign:

The Prospecting Flow:

The AI Quality Control Flow:

The Real Costs: What It Takes to Run This Stack

One of the biggest questions we get: "What does this actually cost?"

Here's the honest breakdown at three different levels:

Beginner Level: $279/month

This is everything you need to get started properly. If you're new to outbound, start here.

What to expect at this level:

• 2,000 leads capacity

• 6,000 emails per month

• 1-5% reply rate typical for most campaigns

• At industry average (1 qualified lead per 250 contacts), you're looking at 8-10 qualified conversations per month if you're doing everything right

• But results vary wildly based on your offer, targeting, and copy quality

Explorer Level: $650/month

You're serious about outbound and want to make a real impact on your pipeline. This is our recommended tier for growing businesses.

What to expect at this level:

• 30,000 leads capacity

• 150,000 emails per month

• Proper volume to test, iterate and scale what works

• At scale, you can run multiple campaigns simultaneously

• With good execution, 5%+ reply rates become achievable

Expert Level: $10,00/year

If you're running this at agency scale like we do, costs jump significantly. But at this point, you're making decisions based on what needs to be done, not what it costs.

What to expect at this level:

• Unlimited sending capacity

• Multiple clients running simultaneously

• Custom workflows for each client

• Advanced AI enrichments at scale

• Results depend entirely on execution quality

What We Learned

Start simpler than you think: You can actually get started with just Apollo if you're testing the waters. Apollo has built-in sequencing, so technically you could run basic campaigns with just that tool + domains + email accounts. But if you want to follow our principles around data quality and deliverability, the Beginner tier is where to start.

Reply rates don't tell the whole story: Most campaigns see 1-5% reply rates. Anything above 5% is exceptional. But positive reply rate (leads actually interested) varies wildly based on your offer, market fit, and targeting. We've seen campaigns with 8% reply rates that generated zero revenue and campaigns with 2% reply rates that closed six figures. Focus on qualified conversations, not vanity metrics.

Tool costs aren't the only expense: A big part of the cost are domains ($10-15 each, you'll need multiple), email hosting, and most importantly your time learning these systems. Budget 2-3 months to get genuinely good at this stack.

Data quality matters more than anything: You can have perfect copy, perfect timing and perfect deliverability, but if you're reaching out to the wrong people or bad email addresses, nothing else matters. This is why we're religious about verification and enrichment.

Don't expect miracles immediately: Outbound is a skill. Our first campaigns were mediocre. Our 50th campaigns were good. Our 200th campaigns were great. The tools accelerate execution, but you still need to learn what works in your specific market.

Why This Stack Actually Works

We didn't pick these tools randomly. Each one plays a clear role:

• Apollo gives us the foundation and reach

• Clay adds context and depth

• Make connects everything and automates the flow

• Smartlead handles delivery and scale

• DeBounce keeps our reputation clean

• GPT + Claude make research and refinement faster

• Notion keeps everything organized

And they compensate for each other's limitations. Apollo's data isn't perfect, so we verify through DeBounce and enrich through Clay. Smartlead is focused on sending, so Make handles all the post-reply automation. Clay is powerful but complex, so we document everything in Notion.

The combination means we can customize for each client with different industries, different outreach styles, and different goals, while maintaining both scale and quality.

The Honest Truth

This stack works for us right now, at our scale, with our clients. Six months ago it looked a little different. Six months from now? It'll probably evolve again.

The tools aren't the point. The system is.

It comes down to knowing exactly what problem each tool solves and not expecting any single one to do everything. We spent months learning what to ignore in each platform. That matters more than knowing all the features.

Access to powerful tools doesn't automatically make you better at outbound. But once you understand how they fit together, that's when you stop spending your days manually verifying email addresses and start actually scaling.